The antibacterial efficacy of azlocillin against Pseudomonas aeruginosa was quantitatively assessed in short-term growth curves in vitro and in a short-term thigh muscle infection in granulocytopenic mice. Ticarcillin was used as the reference drug. The antibacterial effect in vitro was expressed as the difference between the logarithms of the numbers of cfu in the presence and absence of antibiotics (log ratio). The log ratio-time curves could be described by a concentration-dependent parameter, i.e., the inhibition constant (i). According to a standard parallel line bio-assay, based on values of i, azlocillin was 3.80 times more active than ticarcillin (coefficient of variation 27%). Tobramycin in a concentration of 0.19 mg/l, which is not effective in itself, potentiated both drugs by a factor of 3.85 (coefficient of variation 26%). In the in vivo model ticarcillin showed a dose-dependent antibacterial effect, whereas the effect of azlocillin was limited and not dose dependent. The effect of tobramycin in a dose of 4 mg/kg was additive for both penicillins. The results obtained in vivo were less favourable for azlocillin than would be predicted on the basis of the in-vitro results.